Approaches for Power Grid Management and Ancillary Services with DC-AC Micro-grids Comprising of PV Farm and Hybrid Energy Storage System
Analytics
35 views ◎11 downloads ⇓
Abstract
The emergence of the concept of DC grids comes from the fact that there is an increasing number of power grid components that naturally operate with DC power. The combined trend of DC sources and loads in power grids is an important player in the sudden growth of research interest in DC micro-grids and their interplay with the main power grid. Even though DC grids are easy to implement on a small scale and their own, the connection of the DC micro-grid with the AC power grid takes research problems to another level. Innovative approaches, methods, distribution system architectures, and control strategies are required to manage and mitigate the problems of involving DC micro-grid in conjunction with conventional power grids. Through this research work, novel, innovative and comprehensive system architectures and control strategies are proposed to mitigate some of the problems of DC micro-grid like DC-AC micro-grid interplay during steady-state and dynamic conditions, tandem operation of DC, and AC micro-grids for frequency regulation. The dissertation also proposes the concept of a DC ring architecture to emulate a small residential community using a common DC-bus-centric structure. The research work aims to provide a comprehensive DC-AC micro-grid design and control architecture that can provide a platform for testing and validation of steady-state and fault time system performance. The proposed methods in each chapter have been incorporated and validated on standard distribution test systems provided by IEEE like the 13 bus and123 bus systems and hence their analysis is crucial and could serve as a reference for future design and control of DC micro-grids. In this work, first, a new approach for improving grid inertia based on a Photo-Voltaic (PV) farm in conjunction with a fully active Hybrid Energy Storage System(HESS) comprising of battery and ultra-capacitor is proposed. The approach captures grid dynamics using the Point of Common Coupling (PCC) angle measurements (δ) at the inverter terminal. The proposed approach in this dissertation demonstrates an overall improvement in the inertial response of the system by up to 25% compared to the conventional frequency-droop approach. It also records a faster settling time than the frequency-droop approach by up to 47 seconds. The approach has been tested using an error minimization-based Proportional-Integral (PI) control and an error minimization-based Optimal Control (LQR) for frequency regulation. This method utilizes the rate of change of PCC angle (dδdt) which is observed to vary based on the proximity of PCC to the grid dynamic location. It has been observed that the proposed architecture controls the Rate of Change of Frequency (RoCoF) and primary and secondary frequency response without the need for frequency-droop information of the system, thereby ensuring grid stability that utilizes renewable energy-based resources. The second contribution is a novel voltage angle minimization-based approach for grid inertia improvement which further updates and validates the proposed test system and primary control that was used for the previous approach discussed above. The proposed method successfully quantifies and differentiates grid dynam-ics from steady-state conditions by measuring the deviation of voltage angles at the DER interconnection bus (∆δ). A Linear Quadratic Regulator (LQR) controller with a quality function is formulated to minimize (∆δ) during the event of grid dynamics for frequency regulation of the distribution system by optimal dispatch of HESS. The proposed LQR architecture is observed to perform better for inertial support when compared with the conventional frequency-droop approach. Third, a novel architecture that dynamically updates LQR penalty gains based on a combination of recursion and energy storage ramp-rate function is proposed. This helps control the HESS devices separately and not based on their size. The proposed approach is considered as an improvement on the static gain LQR with the added attribute of being robust to uncertainty in the control input. Finally, a concept of DC Ring where 4 DERs are connected to emulate the electrical network of a small residential community is introduced. An Alternating Direction Method of Multipliers (ADMM) based Multi-Input-Multi-Output (MIMO) identification is used to completely identify the dynamic state-space of 4 DERs along with AC-interfacing two 3-phase d−q inverters. This approach is used to generate globally optimized LQR control actions for each of the 4 DERs based on individual weighting factors provided by ADMM. These weighting factors are deduced by drawing a consensus among each output to control the input transfer function. The approach generates globally coordinated control set-points for all DERs that can detect the change in (∆δ) corresponding to the fault event.